Research question and relevance
Question: How do release revisions and the information available at each vintage affect nowcast and forecast quality?
Real-time evaluation matters because today's revised macro series contain information unavailable to a forecaster in the past. The project makes availability dates, vintage labels, transformations, and source limitations explicit.
Original paper or report: This is a project-originated forecasting engine. Any future paper replication must add the exact citation, original sample, specification, and benchmark result before claiming equivalence.
Data sources and sample
Adapters support FRED and ALFRED plus guarded BLS and BEA current-data APIs. Synthetic fixtures allow offline validation. Current BLS and BEA API rows are labeled latest_revised; they are not presented as first-release historical observations.
| Source | Coverage and provenance |
|---|---|
| FRED / ALFRED | Series-specific sample and vintage metadata. |
| BLS / BEA | Current revised public API data under explicit authorization. |
| Fixtures | Synthetic offline samples for schema and pipeline tests. |
Replication methodology
- Record source retrieval time and release availability.
- Normalize units, frequency, adjustment, and transformations.
- Construct vintage-consistent information sets.
- Backtest models without revised-data leakage.
- Publish forecast outputs with source and revision manifests.
Robustness checks cover alternative windows, model classes, transformations, missing-value behavior, retry paths, and unsupported provenance claims.
Original, replicated, and updated results
| Result stage | Status | Interpretation |
|---|---|---|
| Original benchmark | Not asserted | No paper-specific magnitude is claimed without a matched benchmark. |
| Replicated result | Pipeline validated | Tests validate ingestion, provenance, schemas, and guarded source behavior. |
| Updated result | Versioned inputs available | Forecast comparisons await a published result manifest and audited vintage sample. |
Publication-quality charts will be linked when the corresponding benchmark and output manifest pass validation.
Reproduction and citation
Run python -m macro_nowcast.pipeline using the documented configuration. Cite the Observatory release, Dataset revision, project, and original agency sources.
Limitations
An audited historical-vintage archive is required for first-release claims. Numerical comparisons remain incomplete until a result manifest records the sample, vintage policy, configuration, metrics, and checksums.